www.nature.com/scientificreports

OPEN Loss of the transcriptional repressor Rev‑erbα upregulates and proliferation in cultured mouse embryonic fbroblasts Sean P. Gillis1,3, Hongwei Yao1,3, Salu Rizal1, Hajime Maeda1, Julia Chang1 & Phyllis A. Dennery1,2*

The transcriptional repressor Rev-erbα is known to down-regulate fatty acid metabolism and expression. In animal models, disruption of Rev-erbα results in global changes in exercise performance, oxidative capacity, and blood levels. However, the complete extent to which Rev-erbα-mediated transcriptional repression of metabolism impacts cell function remains unknown. We hypothesized that loss of Rev-erbα in a mouse embryonic fbroblast (MEF) model would result in global changes in metabolism. MEFs lacking Rev-erbα exhibited a hypermetabolic phenotype, demonstrating increased levels of and oxidative . Rev-erbα deletion increased expression of II, transketolase, and ribose-5-phosphate isomerase involved in glycolysis and the pentose phosphate pathway (PPP), and these efects were not mediated by the transcriptional activator BMAL1. Upregulation of oxidative phosphorylation was not accompanied by an increase in mitochondrial biogenesis or numbers. Rev-erbα repressed

proliferation via glycolysis, but not the PPP. When treated with ­H2O2, cell viability was reduced in Rev- erbα knockout MEFs, accompanied by increased ratio of oxidized/reduced NADPH, suggesting that perturbation of the PPP reduces capacity to mount an antioxidant defense. These fndings uncover novel mechanisms by which glycolysis and the PPP are modulated through Rev-erbα, and provide new insights into how Rev-erbα impacts proliferation.

Abbreviations 2DG 2-Deoxy-glucose Aldoc DMPO 5,5-Dimethyl-1-pyrroline N-oxide HO-1 Heme oxygenase 1 KO Knockout MEF Mouse embryonic fbroblast NADPH Nicotinamide adenine dinucleotide phosphate PPP Pentose phosphate pathway OT Oxythiamine Pdhb Pyruvate dehydrogenase subunit beta e1 RORE ROR response element RPIA Ribose-5-phosphate isomerase SOD2 Manganese superoxide dismutase 2 Timm23 Translocase of inner mitochondrial membrane 23 TKT Transketolase

1Department of Molecular Biology, Cellular Biology, and Biochemistry, Brown University, 185 Meeting Street, Providence, RI 02912, USA. 2Department of Pediatrics, Warren Alpert Medical School of Brown University, 593 Eddy St Suite 125, Providence, RI 02903, USA. 3These authors contributed equally: Sean P. Gillis and Hongwei Yao. *email: [email protected]

Scientifc Reports | (2021) 11:12356 | https://doi.org/10.1038/s41598-021-91516-5 1 Vol.:(0123456789) www.nature.com/scientificreports/

TOMM20 Translocase of outer mitochondrial membrane 20 WT Wild type

Rev-erbα is a nuclear receptor and transcriptional repressor that forms a key component of the core circadian clock. Rev-erbα represses expression of Arntl, a gene encoding the transcriptional activator BMAL1 through competition for ROR response element (RORE) binding with the transcriptional activator RORα. Tus, Rev- erbα can infuence target gene expression through direct binding of ROREs and RevDR2 motifs or through its indirect efect on BMAL1/CLOCK transcriptional activation via E-Box elements­ 1–5. Tis -translation factor feedback loop regulates expression of a signifcant portion of the genome. Amongst the processes that are controlled by this mechanism are metabolism, the response, and proliferation­ 4,6. In hepatocytes, cells, and fbroblasts, Rev-erbα infuences metabolic gene expression. Changes in expression of metabolic following Rev-erbα perturbation have been shown to impact several pro- cesses such as exercise performance and glucose ­levels7. In excised skeletal muscle, Rev-erbα levels are directly correlated with oxidative capacity, since stabilization of Rev-erbα increases oxygen consumption­ 8,9. However, the extent to which Rev-erbα impacts glycolysis, and whether the efect of Rev-erbα on expression of rate limit- ing glycolytic enzymes is conserved is unclear. Additionally, while datasets detailing transcriptional targets of Rev-erbα in skeletal muscle cells, hepatocytes, and fbroblasts have been developed, there remains a need for mechanistic studies to determine whether these changes have efects on essential cellular functions­ 10–12. Tis study utilized a mouse embryonic fbroblast (MEF) loss or gain of function model to examine the efect of Rev-erbα disruption on global metabolism. In the absence of Rev-erbα, cells demonstrated an increase in glycolysis, oxidative phosphorylation, proliferation, and vulnerability to oxidative stress. Tese phenotypes were accompanied by specifc upregulation of genes involved in glycolysis and the non-oxidative pentose phosphate pathway (PPP). Upregulated expression of these genes may provide a mechanistic basis for increased metabolism and proliferation resulting from perturbation of Rev-erbα. Results With Rev‑erbα disruption, MEFs demonstrate increased levels of glycolysis and oxidative phosphorylation. We previously demonstrated that mouse fbroblasts, which expressed a stabilized phosphomimetic form Rev-erbα resistant to proteasomal degradation (SD), are more resistant to nutrient dep- rivation due to increased oxidative phosphorylation (Oxphos)13. Tus, it was hypothesized that Rev-erbα dis- ruption would manifest the reverse phenotype, namely decreased Oxphos. To test this hypothesis, we utilized immortalized Rev-erbα wild type (WT) and knockout (KO) MEFs, as evidenced by Rev-erbα gene expression (Fig. 1A). We also verifed that Rev-erbα protein levels were not detected in Rev-erbα KO MEFs (Fig. 1B). Mito- chondrial stress tests were then performed using a XF24 Seahorse Analyzer. As shown in Fig. 1C, Rev-erbα KO MEFs demonstrated an increased basal oxygen consumption rate (OCR) compared to WT cells. To examine whether increased oxidative phosphorylation resulted in a bioenergetic shif away from glycolysis, glycolysis stress tests were performed to measure extracellular acidifcation (ECAR). Surprisingly, higher levels of basal glycolysis were observed in Rev-erbα KO MEFs compared to WT cells (Fig. 1D). To confrm the latter, glycolytic rate assays were also performed to remove the confounding efect of mitochondrial acidifcation derived from ­CO2 on extracellular acidifcation. Results were consistent with those of the glycolytic stress test, with Rev-erbα KO MEFs demonstrating increased basal and compensatory glycolysis compared to WT (Fig. 1E). Overall, the increased Oxphos and glycolysis in the MEF KO demonstrate that upon Rev-erbα disruption, there is global metabolic activation.

Rev‑erbα KO MEFs do not exhibit increased mitochondrial mass or mtDNA biogenesis. In a Rev-erbα stabilized cell line, increased oxidative phosphorylation was associated with an increase in mitochon- drial area and ­biogenesis13. Additional studies in skeletal muscle also indicate that Rev-erbα overexpression can increase mitochondrial biogenesis­ 8,9. To interrogate whether increased Oxphos in the Rev-erbα KO MEFs was associated with an increase in mitochondrial mass, expression of the protein translocase of inner mitochondrial membrane 23 (Timm23) was measured. Rev-erbα KO MEFs did not exhibit increased Timm23 gene expression relative to WT (Supplemental Fig. 1A). Levels of the protein translocase of outer mitochondrial membrane 20 (TOMM20) was also unchanged following Rev-erbα disruption (Supplemental Fig. 1B). In addition, we meas- ured mitochondrial DNA copy number, and quantifed the mtDNA/nDNA ratio by qPCR. As shown in Sup- plemental Fig. 1C and D, the Ct values of 16S, ND1 and HKII as well as ratio of 16S/HKII and ND1/HKII DNA were not altered in Rev-erbα KO MEFs compared to WT cells. Tese results suggest that the hypermetabolic state of Rev-erbα KO MEFs is not due to increased mitochondrial mass.

Expression of glycolytic and non‑oxidative PPP enzymes is upregulated following Rev‑erbα disruption. To elucidate whether the loss of Rev-erbα mediated transcriptional repression of specifc genes regulating glycolysis, a glycolysis gene array was performed. Tis array included genes encoding enzymes involved in glycolysis, gluconeogenesis, and the PPP (Table 1). In Rev-erbα KO MEFs two of the genes most highly upregulated relative to WT were ribose-5-phosphate isomerase (RPIA) and transketolase (TKT) of the non-oxidative PPP. Tese are involved in the production of ribose-5 phosphate needed for nucleotide synthesis and the shuttling of metabolic intermediates for glycolysis, ­respectively14,15. Other targets upregulated upon Rev- erbα disruption included aldolase c, an isoform in the canonical glycolysis pathway­ 16,17, and pyruvate dehydro- genase subunit b e1 (pdhb1), one of the genes encoding a subunit of the pyruvate dehydrogenase complex­ 18,19 (Table 1). Quantitative polymerase chain reaction (qPCR) verifed that Rev-erbα KO MEFs had an approxi- mately two-fold increase in each gene (i.e., RPIA, TKT and pdhb1) compared to WT (Fig. 2). Serum starvation

Scientifc Reports | (2021) 11:12356 | https://doi.org/10.1038/s41598-021-91516-5 2 Vol:.(1234567890) www.nature.com/scientificreports/

Figure 1. Global metabolism is increased upon loss of Rev-erbα. (A) Rev-erbα gene expression in Rev-erbα KO and WT MEFs. n.d. denotes not detected. (B) Western blot analysis to measure Rev-erbα protein levels in WT and Rev-erbα KO MEFs. β-actin was used as a loading control. Densitometry analysis of Rev-erbα levels afer normalization to β-actin in Rev-erbα KO and WT MEFs. Minimum of n = 5 independent experiments. n.d. denotes not detected. (C) Oxygen consumption rate (OCR) trace from the Seahorse mitochondrial stress test afer sequential injection of oligomycin (Oligo), FCCP and antimycin A/rotenone (Anti/Rot). Levels of basal respiration calculated in Rev-erbα KO and WT MEFs. (D) Extracellular acidifcation rate (ECAR) trace from glycolysis stress tests afer sequential injection of glucose (Glu), oligomycin (Oligo) and 2-deoxy-d-glucose (2-DG). Levels of basal glycolysis calculated in WT and Rev-erbα KO cells. (E) Proton efux rate measured from Seahorse glycolytic rate assays. Levels of basal and compensatory glycolysis were calculated in WT and Rev-erbα KO MEFs. All Seahorse assay traces were normalized to 10,000 cells. Seahorse calculations represented as a ratio to WT group. N = 3–4 independent experiments. Error bars represented as mean ± SEM. *p < 0.05, ***p < 0.001 versus WT.

can synchronize cells into G0 quiescence. As shown in Supplemental Fig. 2A, serum deprivation signifcantly increased Rev-erbα protein levels in WT cells. Interestingly, RIPA gene expression was increased whereas TKT transcription was reduced in synchronized cells compared to unsynchronized cells (Supplemental Fig. 2B). As with unsynchronized cells, increased expression of RPIA, TKT, and pdhb1 following Rev-erbα disruption was also observed when the cells were synchronized to control for nutrient availability (Supplemental Fig. 2C). Expression of rate limiting enzymes in glycolysis were additionally measured. In fact, hexokinase II gene expres- sion was elevated in Rev-erbα KO MEFs relative to WT cells, and this was maintained following synchronization (Fig. 2; Supplemental Fig. 2B). Overall, these data demonstrate a cell-cycle specifc Rev-erbα expression and its repressive role in the non-oxidative PPP and in glycolysis.

Inverse efects of Rev‑erbα disruption and stabilization on glycolysis and non‑oxidative PPP expression. Fibroblasts expressing a phosphomimic form of Rev-erbα (SD) to protect Rev-erbα from pro- teasomal degradation and extent protein half-life, were utilized­ 20. Tese cells demonstrated a threefold increase in Rev-erbα protein levels relative to WT (Fig. 3A) and reduced levels of basal glycolysis as expected (Fig. 3B). Te phosphomimic form of Rev-erbα migrated more slowly on the SDS gel, most likely owing to reduced inter- actions with sodium dodecyl sulfate or negatively charged aspartate­ 21,22. However, these cells maintained a high proton efux following administration of 2-deoxy-glucose (2-DG) compared to WT cells. Tis led to reduction in basal glycolysis in SD cells compared to WT cells (Fig. 3B).

Scientifc Reports | (2021) 11:12356 | https://doi.org/10.1038/s41598-021-91516-5 3 Vol.:(0123456789) www.nature.com/scientificreports/

Genes Fold change Full name of genes Aldob − 3.50 , fructose-bisphosphate Aldoc 2.11 Aldolase C, fructose-bisphosphate Cs 1.67 Citrate synthase Dlat 1.5 Dihydrolipoamide S-acetyltransferase (E2 component of pyruvate dehydrogenase complex) Dlst − 8.54 Dihydrolipoamide S-succinyltransferase (E2 component of 2-oxo-glutarate complex) Eno1 − 7.20 1, alpha non-neuron Eno2 − 1.66 Enolase 2, gamma neuronal Fbp1 − 1.31 Fructose bisphosphatase 1 G6pc3 − 5.12 Glucose 6 phosphatase, catalytic, 3 G6pdx − 1.36 Glucose-6-phosphate dehydrogenase X-linked Galm − 1.19 Galactose mutarotase Gpi1 1.39 Glucose phosphate isomerase 1 Gys1 − 1.76 Glycogen synthase 1, muscle Gys2 − 1.62 Glycogen synthase 2 Hk2 − 1.76 Hexokinase 2 Hk3 − 1.72 Hexokinase 3 Idh3a 1.02 Isocitrate dehydrogenase 3 (NAD +) alpha Idh3b − 2.00 Isocitrate dehydrogenase 3 (NAD +) beta Idh3g − 1.36 Isocitrate dehydrogenase 3 (NAD +), gamma Mdh1 1.1 Malate dehydrogenase 1, NAD (soluble) Ogdh 1.75 Oxoglutarate dehydrogenase (lipoamide) Pcx 1.25 Pdha1 − 1.9 Pyruvate dehydrogenase E1 alpha 1 Pdhb 4.82 Pyruvate dehydrogenase (lipoamide) beta Pdk1 1.79 Pyruvate dehydrogenase kinase, isoenzyme 1 Pdk2 − 16.05 Pyruvate dehydrogenase kinase, isoenzyme 2 Pdk3 − 1.77 Pyruvate dehydrogenase kinase, isoenzyme 3 Pdp2 − 1.12 Pyruvate dehyrogenase phosphatase catalytic subunit 2 Pgam2 − 2.66 2 Pgk2 − 4.51 2 Phkb − 3.58 Phosphorylase kinase beta Phkg1 − 2.15 Phosphorylase kinase gamma 1 Prps1 1.55 Phosphoribosyl pyrophosphate synthetase 1 Prps1l1 − 1.36 Phosphoribosyl pyrophosphate synthetase 1-like 1 Prps2 1.61 Phosphoribosyl pyrophosphate synthetase 2 Pygl 170.28 glycogen phosphorylase Pygm − 2.18 Muscle glycogen phosphorylase Rbks 1.78 Ribokinase Rpe − 1.9 Ribulose-5-phosphate-3-epimerase Rpia 1.96 Ribose 5-phosphate isomerase A Sdhb 1.39 Succinate dehydrogenase complex, subunit B, iron sulfur (Ip) Sucla2 1.00 Succinate-Coenzyme A ligase, ADP-forming, beta subunit Suclg2 − 2.11 Succinate-Coenzyme A ligase, GDP-forming, beta subunit Taldo1 − 1.79 Transaldolase 1 Tkt 2.04 Transketolase

Table 1. Changes in gene expression in Rev-erbα KO MEFs relative to WT by a glycolysis gene array.

Although Oxphos was increased, fuel fexibility tests revealed that following Rev-erbα disruption, there was a reduced dependence on glutamine and fatty acids for energy production, while glucose utilization remained consistent (Fig. 3C). Tese results suggest anaplerotic reactions may contribute to increased Oxphos in Rev-erbα KO cells. Expression of RPIA and TKT was reduced in Rev-erbα SD cells relative to WT (Fig. 3D), suggesting that this inverse relationship may extend to non-oxidative PPP enzymes. Expression of the Pdhb1 gene was addition- ally reduced in the Rev-erbα SD cells (Fig. 3D), which confrmed negative regulation of this gene by Rev-erbα.

Rev‑erbα does not regulate expression of target genes through BMAL1. To determine whether Rev-erbα modulates metabolic gene expression through BMAL1, we frst determined expression of the Arntl gene in Rev-erbα KO cells. As shown in Supplemental Fig. 3A, Rev-erbα KO MEFs exhibited increased expres-

Scientifc Reports | (2021) 11:12356 | https://doi.org/10.1038/s41598-021-91516-5 4 Vol:.(1234567890) www.nature.com/scientificreports/

Figure 2. Loss of Rev-erbα leads to upregulation of genes encoding enzymes in the glycolytic and non- oxidative pentose phosphate pathways. Expression of RPIA, TKT, pdhb1, and hexokinase II genes in MEFs cultured in medium containing 10% FBS. n = 5 independent experiments. Error bars represented as mean ± SEM. *p < 0.05, *p < 0.01 versus WT.

Figure 3. Rev-erbα stabilization reduces levels of glycolysis as well as glycolytic and PPP expression. (A) Western Blot analysis to measure Rev-erbα protein levels in Rev-erbα KO, SD, and WT cells. β-Actin was used as a loading control. Densitometry analysis of Rev-erbα levels in Rev-erbα KO, SD, and WT cells. n.d. denotes not detected. (B) Proton efux rate trace for WT and Rev-erbα SD MEFs normalized to cell number. Anti/Rot: antimycin A/rotenone. Levels of basal and compensatory glycolysis represented as ratios to WT. (C) Fuel dependency analysis in WT and Rev-erbα KO MEFs. Glutamine, fatty acid, and glucose dependency are denoted as a percentage of total fuel oxidation. (D) Expression of TKT, RPIA, and pdhb1 genes in WT and Rev-erbα SD cells. n = 3 independent experiments. Error bars represented as mean ± SEM. *p < 0.05, **p < 0.01, ***p < 0.001 versus WT.

sion of Arntl relative to WT cells. Te promoter of RPIA contains E-Box elements which could potentially bind ­BMAL11–3, allowing for an indirect mechanism for the observed increased RPIA expression in Rev-erbα KO MEFs. To test this hypothesis, Arntl expression was partially ablated using small interfering RNAs (Supplemen- tal Fig. 3B). Tis only led to a modest decrease in RPIA gene expression in Rev-erbα KO cells (Supplemental Fig. 3B). Interestingly, upon robust overexpression of Arntl in WT cells, expression of hexokinase II, RPIA, TKT, and pdhb1 were unchanged (Supplemental Fig. 3C and D). Tus, Rev-erbα is unlikely to indirectly control the expression of these glycolysis and PPP enzymes through BMAL1.

Scientifc Reports | (2021) 11:12356 | https://doi.org/10.1038/s41598-021-91516-5 5 Vol.:(0123456789) www.nature.com/scientificreports/

Figure 4. Loss of Rev-erbα increases proliferation, cell growth, and vulnerability to oxidative stress. (A) EdU incorporation assays performed with WT and Rev-erbα KO MEFs. (B) A growth curve of WT and Rev-erbα KO MEFs with counts over a 5-day span at daily intervals. (C) Proliferation assays performed with WT and Rev- erbα SD MEFs. (D) A growth curve of WT and SD MEFs with counts over a 4-day span at daily intervals (right panel). (E) Basal glycolysis of Rev-erbα KO and WT MEFs treated with oxythiamine (OT) or 2-deoxy-glucose (2-DG) for 24 h. Glycolytic levels represented as a ratio to WT/vehicle group. (F) Proliferation assays performed with Rev-erbα KO and WT MEFs following treatment with 0.5 mM of OT and 2-DG, or transfection with RPIA siRNA. (G) Rev-erbα protein was measured by Western blot in WT MEFs treated with H­ 2O2 (500 µM) for 24 h. (H) Viability of WT and Rev-erbα KO MEFs following 24 h treatment with 500 µM H­ 2O2 measured by trypan blue exclusion. (I) Ratio of NADP­ +/NADPH in WT and Rev-erbα KO MEFs. n = 3 independent experiments. Error bars represented as mean ± SEM. *p < 0.05, **p < 0.01, ***p < 0.001 versus WT (A–D, H and I), vehicle (E–G) by unpaired t test; †p < 0.05 versus WT/2-DG group.

Cell numbers and proliferation are increased following Rev‑erbα disruption in MEFs. Te enzymes RPIA and TKT are known to be important for the synthesis of ribose-5-phosphate, a precursor for nucleotide ­biosynthesis23–25. It may be possible that increased nucleotide and metabolism in Rev- erbα KO MEFs facilitate the demands of increased proliferation. To determine if Rev-erbα impacts cell prolif- eration, a 5-ethynyl-2′-deoxyuridine (EdU) incorporation assay was performed. Rev-erbα KO MEFs demon- strated increased proliferation relative to WT (Fig. 4A). To determine if the increased proliferation manifested as increased cell numbers, a 5-day growth curve as performed. Rev-erbα KO MEFs demonstrated dramatically increased cell numbers by day 5 (Fig. 4B). Conversely, proliferation and cell numbers were reduced in the Rev- erbα SD MEFs compared to WT (Fig. 4C, D). To determine whether increased proliferation was dependent on the PPP based nucleotide biosynthesis or glycolysis, Rev-erbα KO and WT cells were treated with the TKT inhibitor ­oxythiamine26–28 or the hexokinase inhibitor 2-DG29,30. Glycolytic rate assays demonstrated that both 2-DG and oxythiamine reduced glycolysis in WT cells (Fig. 4E). Tese efects were attenuated in Rev-erbα KO cells (Fig. 4E). Furthermore, treatment with 2-DG, but not oxythiamine, reduced cell proliferation in both Rev-erbα KO and WT MEFs (Fig. 4F). Similarly, EdU incorporation was not altered in Rev-erbα KO transfected with RPIA siRNA transfection in the presence or absence of oxythiamine compared to the scrambled siRNA group (Fig. 4F). Compared to WT cells, 2-DG- mediated reduction of EdU incorporation was less in Rev-erbα KO cells (Fig. 4F). Tese results demonstrate that increased glycolysis contributes to the hyperproliferative phenotype of Rev-erbα KO MEFs compared to WT cells.

Loss of Rev‑erbα increases vulnerability to oxidative stress. Our previous studies showed that sta- bilization of Rev-erbα in MEFs is protective against hydrogen peroxide (H­ 2O2)-induced oxidative stress and that oxidative stress can infuence Rev-erbα transcriptional ­repression13,31. At baseline, Rev-erbα KO and WT MEFs exhibited comparable levels of oxidative stress, as measured by 5,5-dimethyl-1-pyrroline N-oxide (DMPO) pro- tein ­adducts32,33 (Supplemental Fig. 4A). Additionally, expression of heme oxygenase 1 (HMOX1) and man- ganese superoxide dismutase 2 (SOD2) genes, encoding enzymes known to upregulated in response to oxida- tive stress, were unchanged upon Rev-erbα ­disruption34–36 (Supplemental Fig. 4B). Furthermore, SOD2 protein

Scientifc Reports | (2021) 11:12356 | https://doi.org/10.1038/s41598-021-91516-5 6 Vol:.(1234567890) www.nature.com/scientificreports/

levels were also unchanged in Rev-erbα KO MEFs (Supplemental Fig. 4C). Although 24 h of ­H2O2 incubation increased Rev-erbα protein levels in WT cells, Rev-erbα KO MEFs exhibited reduced viability relative to WT cells (Fig. 4G, H). Rev-erbα stabilization has been shown to upregulate HMOX1 and SOD2 gene expression in 13 response to H­ 2O2 incubation­ . Tis fnding led us to hypothesize that expression of these enzymes would be reduced following Rev-erbα disruption in response to oxidative stress. Surprisingly, HMOX1 expression was increased in H­ 2O2-treated KO MEFs, whereas SOD2 was unchanged, compared to WT cells (Supplemental Fig. 4D). Tis suggests that, in contrast to the Rev-erbα SD, where increased expression of antioxidant enzymes increased cell ­viability13, increased expression of HO-1 does not infuence viability of Rev-erbα KO. Although Rev-erbα KO MEFs exhibited increased expression of TKT and RPIA, inhibition of these enzymes did not afect proliferation (Fig. 4F, right panel). Te non-oxidative PPP is downstream of the oxidative PPP, which produces nicotinamide adenine dinucleotide phosphate (NADPH) necessary for the reduction of the key antioxidant enzyme glutathione. Rev-erbα KO MEFs exhibited a higher ratio of NADP­ +/NADPH compared to WT cells (Fig. 4I), suggesting that perturbation of the PPP following Rev-erbα disruption may underlie reduced viability to oxidative stress. Discussion Tis study demonstrates that Rev-erbα regulates glycolysis by specifc upregulation of enzymes in the PPP as well hexokinase, a rate-limiting enzyme in glycolysis. Tis is associated with increased proliferation and growth in MEFs. Several reports have linked Rev-erbα to oxidative ­phosphorylation9,13,37. Here we show that Rev-erbα KO MEFs had increased oxidative phosphorylation relative to WT cells. Paradoxically, Rev-erbα SD MEFs also demonstrated an increase in oxidative phosphorylation. Previous studies indicate overexpression of Rev-erbα in fbroblasts and skeletal muscle cells leads to an increase in mitochondrial mass and biogenesis relative to ­WT8,9,13, we did not see any change levels of mitochondrial mass and mtDNA content, suggesting that loss of Rev-erbα does not result in changes in mitochondrial mass or numbers. Tis diference in phenotype may explain the paradoxical relationship between Rev-erbα levels and oxidative phosphorylation. We noticed that HKII mRNA was increased in KO cells, there were no changes of HKII DNA copy number in these cells compared to WT cells. Further study is required to determine the mechanisms underlying discrepancies between DNA copy number and mRNA levels in Rev-erbα KO cells. It remains elusive whether Rev-erbα deletion upregulates metabolism and proliferation in other types of cells, including epithelial cells, muscle cells and hepatocytes. While Rev-erbα levels can govern the expression of genes involved in hepatocyte cholesterol ­metabolism38,39, gluconeogenesis­ 40–42, and the immune response­ 43–45, studies detailing the efect of Rev-erbα levels on expression of glycolytic enzymes have been less conclusive, with diferent patterns of expression of rate limiting enzymes evident in skeletal muscle cells, fbroblasts, and ­hepatocytes7,13. Additionally, little research has been done to interrogate the efect of Rev-erbα on the overall output of the glycolytic pathway. Rev-erbα disruption in MEFs led to specifc upregulation of hexokinase, TKT, pdhb1, and aldolase c in unsynchronized cells. Te presence of an E-Box element in the promoter of RPIA gene suggested there may be an indirect mechanism of transcriptional regulation through BMAL1 which known to have its expression repressed by Rev-erbα1–3. Additionally, recent evidence has suggested that expression of aldolase c may be driven in part through BMAL1­ 46. However, we found that while expression RPIA, TKT, and pdhb1 demonstrated an inverse relationship with Rev-erbα levels, overexpression of BMAL1 in WT MEFs did not lead to changes in expression of these genes. Tis suggest that BMAL1 may not be involved in upregulation of these genes in Rev-erbα KO cells. Further experiments with BMAL1 deletion or overexpression in both Rev-erbα KO and WT MEFs in the same circadian phase would reveal whether BMAL1 mediates hypermetabolism and hyperproliferation in Rev-erbα KO cells. It is possible that Rev-erbα exerts control of glycolysis and the PPP through a mechanism that is divergent from the traditional circadian clock. Amongst the more notable studies providing evidence for this was the fnding that Rev-erbα can interact with the liver lineage hepatocyte nuclear factor 6 to repress gene expression in ­hepatocytes11. A recent study demonstrates that Rev-erbα controls more extensive transcription in the liver under conditions of metabolic perturbation such as mis-timed feeding­ 47. Tis suggests that the repressive action of Rev-erbα does not drive basal rhythmicity in metabolic activity but serves to bufer against metabolic perturbation in the liver. Recent evidence in a Rev-erb α/β double KO MEF cell line found enrichment for binding of the hepatocyte nuclear factor 1 and nuclear transcription factor 1 in the promoters of genes that halted circadian oscillation following Rev-erb disruption. While our experiments in synchronized MEFs suggest that RPIA, TKT, pdhb1, and hexokinase II may be under circadian control, these genes were not found to be oscillating in double KO ­MEFs12. Future work is required to determine the extent to which Rev-erbα mechanistically infuences the expression of these metabolic genes and the circadian genes, including Cry, Per and Nfl3, in a clock-dependent or independent manner. Cells become quiescent (entering the G0 phase) dur- ing serum deprivation. Interestingly, Rev-erbα was increased in synchronized MEFs with serum deprivation. Tis suggests that Rev-erbα is regulated in a cell-cycle dependent manner. Serum shock is able to phase-lock the cells and alters Rev-erbα expression in ­fbroblasts48. Future study using serum shock (i.e., treating cells with high concentrations of serum) will further validate the role of Rev-erbα in modulating glycolysis, the PPP and proliferation. Glycolysis and the PPP make important contributions toward cellular proliferation. cells are known to increase aerobic glycolysis via the Warburg efect to drive uncontrolled ­proliferation49–51. A number of studies have suggested that Rev-erbα may be a viable therapeutic target for the treatment of breast, colon, brain, and gastric ­cancers52–57. Studies in cancerous and non-cancerous cell lines have demonstrated that incubation with the synthetic Rev-erbα agonists SR9009 and GSK4112 can impede ­proliferation52,54,58. However, recent fndings have indicated that SR9009 may have nonspecifc efects on cell numbers and delay cell cycle progression in the absence of Rev-erbα37. Te fndings of the present study indicate that Rev-erbα KO and SD cells have increased

Scientifc Reports | (2021) 11:12356 | https://doi.org/10.1038/s41598-021-91516-5 7 Vol.:(0123456789) www.nature.com/scientificreports/

and decrease proliferation relative to WT cells, respectively, which corroborates with previous research suggest- ing that Rev-erbα can regulate proliferation. We show that Rev-erbα KO MEFs displayed increased glycolysis and proliferation resembling the Warburg efect. Tis hyperproliferative phenotype may be driven by upregulation of the non-oxidative PPP. Te PPP is known to be the main source of nucleotides needed for DNA replication during cell division, with the non- oxidative arm of the PPP being predominantly involved in the production of ribose-5 via RPIA, which directly catalyzes the production of ribose-5. Expression of these enzymes is important to proliferation of cancer cell lines. In cell lines, silencing of TKT led to cell cycle arrest coupled with metabolic fux towards glycolysis away from nucleotide biosynthesis­ 23. Additionally, in an in vitro model of colon cancer cells, silenc- ing of RPIA led to reduced nucleotide incorporation and cell survival­ 24. Although we could demonstrate a clear inverse relationship between Rev-erbα levels and non-oxidative PPP gene expression, we were unable to observe any associated change in cell proliferation following inhibition of TKT or ablation of RPIA. In contrast, inhibition of hexokinase II reduced proliferation. Elevated expression of hexokinase II is known to be a hallmark of several cancer cell lines providing needed metabolic intermediates for glycolysis, oxidative phosphorylation, and nucleotide ­biosynthesis59–61. In WT MEFs, inhibition of hexokinase II using 2-DG led to a reduction in nucleoside incorporation. Tese efects were attenuated in Rev-erbα KO cells, which may be due to high glycolysis in these cells. Increased glycolysis, characteristic of the Warburg efect, has also been shown to be driven by phosphofructokinase and , which mediate additional rate limiting steps in ­glycolysis62–65. Expression of these enzymes was unchanged following Rev-erbα disruption, suggesting hexokinase alone drove increased glycolysis and proliferation in this background. Tis insight provides new evidence that loss of transcriptional repression not associated with oncogenesis may drive a hypermetabolic, hyperprolifera- tive phenotype. While perturbation of the non-oxidative PPP did not afect glycolysis or proliferation, it is possible that Rev- erbα-mediated changes in non-oxidative PPP expression may afect antioxidant defense. Oxidative stress has been previously demonstrated to drive Rev-erbα transcription through nuclear erythroid factor 2­ 31. Conversely, to date, the efect of Rev-erbα on the oxidative stress response has been controversial. Rev-erbα stabilization provides protection in fbroblasts against hydrogen peroxide stress through induction of HO-1, SOD2, and ­catalase13. However, in a breast cancer cell line, Rev-erbα inhibits Poly [ADP-ribose] polymerase 1 leading to an increase 56 in DNA damage and cell death­ . Te increase in Rev-erbα in WT cells treated with H­ 2O2 may be compensatory and protective against cell death. Rev-erbα KO MEFs under oxidative stress exhibited reduced viability that was not accompanied by a reduction in HO-1 or SOD2 expression. Tis suggests that the diminished cellular pool of NADPH following Rev-erbα disruption contributes more to increased oxidative vulnerability than antioxi- dant enzymatic expression or DNA damage. Te possibility that diminished NADPH contributes to oxidation of the important antioxidant GSH in the Rev-erbα KO cells remains to be determined. Te PPP is thought to contribute to the regeneration of the reducing agent NADPH. Although Rev-erbα KO cells displayed increased glycolysis and PPP, levels of NADPH were reduced in these cells. We surmise that NADPH may be consumed for macromolecule (e.g. fatty acids) synthesis to increase proliferation in the Rev-erbα KO cells. Further studies using a metabolic fux assay are required to determine whether Rev-erbα deletion causes a metabolic switch between glycolysis and the PPP. In summary, disruption of Rev-erbα results in increased glycolysis, proliferation, and cell growth. Tese phenotypic changes are driven by specifc changes in hexokinase II expression that is inversely proportional to Rev-erbα levels. Furthermore, specifc enzymes of the PPP are also increased in response to Rev-erb disruption, leading to diminished pools of NADPH and subsequent reduced protection against oxidative stress. Experimental procedures Cell culture, synchronization and treatment. MEFs were used as an in vitro model for this study. Fibroblasts were prepared from E13.5 WT and Rev-erbα Global Knockout Embryos (gif of Mitch Lazar, Univer- sity of Pennsylvania, Philadelphia, PA, USA) using standard protocols. Te cells were transfected with an SV40 plasmid for immortalization and maintained in DMEM (Caisson Laboratories, Smithfeld, UT, USA) supple- mented with 10% bovine serum, penicillin/streptomycin, and glutamine. Cells were maintained at 37 °C under 5% ­CO2 condition. Synchronized MEF cells were seeded, allowed to attach for 2 h, afer which they were cultured in serum free DMEM for 24 h. Cells were then cultured in serum containing DMEM for an additional 48 h. Serum starvation followed by replacement with serum containing medium aligns the cells in G0 quiescence. For transketolase inhibition experiments, cells were treated with oxythiamine (Sigma Aldrich, catalog#: 614-05-1) for 24 h. Cells were treated and collected for endpoint measurements at similar times of day for consistency.

Mitochondrial stress test. Parameters of oxidative phosphorylation were measured using the mitochon- drial stress test on the Seahorse XF24 Bioanalyzer (Agilent, Santa Clara, CA, USA) as described previously­ 66. Cells were seeded in Seahorse assay plates 24 h in air (21% O­ 2/5% ­CO2) prior to assay. Cells were inspected the day of the assay to ensure uniform confuence and one well of each genotype was sacrifced for cell counting. OCR was normalized to 10,000 cells per well and all parameters were normalized to the average of WT group as a ratio. Basal respiration is defned as the initial level of oxygen consumption rate, and maximal respiration is the level of oxygen consumption rate measured following FCCP injection.

Glycolysis stress test. Glycolysis stress tests were performed as an indicator of glycolytic levels as described previously­ 67. MEF cells were seeded and incubated overnight as described for mitochondrial stress tests. ECAR was used as an index of glycolysis with initial levels normalized to cell number for each genotype. Basal glycolysis

Scientifc Reports | (2021) 11:12356 | https://doi.org/10.1038/s41598-021-91516-5 8 Vol:.(1234567890) www.nature.com/scientificreports/

was calculated based on the diferences in ECAR afer glucose injection, and the data are expressed as a relative ratio of 1 normalized to WT.

Glycolytic rate assay. MEF cells were seeded and incubated overnight as described for mitochondrial stress tests. Proton efux rate response following injections of the electron transport chain inhibitors rotenone/ antimycin A and the glucose analog 2-DG was measured to determine basal and compensatory glycolysis as described ­previously67. Data were analyzed using the Seahorse Glycolytic Rate assay report generator, and basal glycolysis was calculated based on the diferences in PER afer 2-DG injection. Te data were normalized to cell number for each genotype, and converted to a ratio of 1 with respect to WT.

Fuel fexibility assay. To determine dependency on glutamine, fatty acids, and glucose, fuel fexibility assays were performed­ 66. MEF cells were seeded and incubated overnight as described for the mitochondrial stress test. Te drop in basal consumption following injections of BEPTES, UK5099, and etomoxir were quanti- fed as a percentage of basal respiration for glutamine, glucose, and fatty acids, respectively.

Glycolysis gene array. A Qiagen Glucose Metabolism ­RT2 Profler PCR Array (Qiagen, Hilden, Germany) was used to determine expression levels of genes involved in glycolysis, gluconeogenesis, or the pentose phos- phate pathway. Cells were pooled from 3 samples in each condition. RNA was isolated using an RNeasy prep kit that was also from Qiagen. Expression of microchip array targets was measured using a GeneChip Instrument (Afymetrix, Santa Clara, CA, USA).

Quantitative RT‑PCR. Quantitative RT-PCR was used to measure relative expression of metabolic enzymes. Relative expression of all mRNA targets was tested using the TaqMan gene expression system (Applied Biosystems, Foster City, CA, USA). Total RNA was extracted using the TRIzol Plus RNA Purifcation Sys- tem (Invitrogen, Carlsbad, CA, USA) or a Qiagen RNeasy Kit (Qiagen, Hilden, Germany). A NanoDrop One (Termo Scientifc, Waltham, MA, USA) instrument was used to test the quantity and concentration of RNA. For arntl silencing and overexpression, cell lysis was performed in plate with a Qiagen RNeasy kit (Germantown, MD) used for downstream RNA processing according to manufacturer protocols. cDNA was synthesized using the TaqMan Reverse Transcription System and run on an Applied Biosystems 7300 Real-Time PCR instrument. TaqMan probes were as follows: nr1d1 (Mm00520711), RPIA (Mm00485790), transketolase (Mm00447559), arntl (Mm00500226), pyruvate kinase muscle isoform (m1/m2) (Mm00834102), pyruvate kinase liver/red blood cell isoform (Mm00443090), phosphofructokinase (Mm01309576), hexokinase I (Mm00439344), hexokinase II (Mm00443385), aldolase c (Mm01298116), and pdhb (Mm00499323). Samples were denatured at 95 °C for 10 min, followed by 40 cycles of denaturing at 95° for 15 s followed by extension at 60 °C for 1 min.

Mitochondrial DNA measurement. DNA was extracted using an organic phenol extraction bufer (phenol:chloroform at a ratio of 5:3), and precipitated with absolute alcohol cooled to − 20 °C containing sodium acetate (0.3 M) as described ­previously66. DNA (40 ng) was used for quantitative PCR to measure the relative copy number of mtDNA (16S and ND1) and nDNA (hexokinase 2, HK2). Te primers are shown as follows, 16S: fwd: 5′-CCG​CAA​GGG​AAA​GAT​GAA​AGAC-3′; rev: 5′-TCG​TTT​GGT​TTC​GGG​GTT​TC-3′, ND1: fwd: 5′-CTA​ GCA​GAA​ACA​AAC​CGG​GC-3′; rev: 5′-CCG​GCT​GCG​TAT​TCT​ACG​TT-3′, HK2: fwd: 5′-GCC​AGC​CTC​TCC​ TGA​TTT​TAG​TGT​-3′; rev: 5′-GGG​AAC​ACA​AAA​GAC​CTC​TTC​TGG​-3′. PCR reaction mixtures included SYBR Green Supermix (Cat#:172-5271, Bio-Rad), primer master mixtures, DNA (20 ng/µl) and free water. Te reactions were performed at 50 °C for 2 min, and following 40 cycles at 98 °C for 10 min, 98 °C for 15 s, 60 °C for 60 s. Analysis of mtDNA/nDNA ratio was be calculated by the classical 2­ -ΔΔCt method used for qPCR analysis.

Gene silencing and overexpression. Lipid mediated siRNA transfection was used to knock down arntl expression. For anrtl silencing, pooled arntl Qiagen fexitube siRNAs (catalog # 1027416) were transfected into KO MEFs using Termo Fisher RNAiMAX reagent with transfection performed per manufacturer instructions. Termo Fisher Ambion siRNA (catalog # AM16708) was used for ribose-5-phosphate isomerase (RPIA) silenc- ing. Arntl was overexpressed in WT MEFs using an Origene pCMV6 vector plasmid (catalog # MR209553) using Termo Fisher Lipofectamine 3000 reagent. RNA was harvested 24 h post transfection.

EdU incorporation assay. Nucleoside incorporation assays were performed as an index of proliferation as described ­previously67. Cells were seeded 24 h prior to assay start time. EdU is a thymidine analog that incor- porated into cell during DNA synthesis in S phase of the cell cycle to quantify the proportion of cell undergoing activate proliferation. Cells were then treated with 3 µM EdU provided by the Click-iT EdU proliferation assay (Invitrogen, Carlsbad, CA, USA). Cells prepared according to manufacturer’s protocol. Flow Cytometry was then performed using a BD FACSAria IIIu at the Brown Flow Cytometry core (Providence, RI, USA), with cells from the main population registering as Alexa Fluor 488 positive deemed proliferative. A minimum of 20,000 events were recorded per sample. For non-oxidative PPP experiments, cells were treated with oxythiamine or transfected with RPIA siRNA 24 h prior to EdU treatment.

Growth curves. For these experiments, 28,000 MEFs were seeded in 6 ­cm2 plates for the WT and Rev-erbα SD genotypes. For Rev-erbα KO cells, we seeded 14,000 cells in a plate so as to avoid over confuence at the end of the experiment since these cells grow signifcantly faster than WT. Tree plates of MEFs were trypinisized and

Scientifc Reports | (2021) 11:12356 | https://doi.org/10.1038/s41598-021-91516-5 9 Vol.:(0123456789) www.nature.com/scientificreports/

counted on a hemocytometer using trypan blue reagent over a period of four to fve days, with live cell counts averaged and standard error of the mean calculated.

Immunoblotting. Western blotting was used to measure relative protein levels. Total protein was isolated from Rev-erbα WT, KO, and SD MEFs. Lysate was mixed with β-mercapaethanol, boiled, run on a 4–12% Sodium Dodecyl Sulfate Polyacrylamide Gel (Invitrogen, Carlsbad, CA, USA), and transferred to a PVDF mem- brane. Membranes were incubated with the following primary antibodies and dilutions: Rev-erbα (#13,418, Cell Signaling Technologies, Danvers, MA monoclonal rabbit antibody, 1:1000), TOMM20 (#42406S Cell Signaling Technologies, Danvers, MA, USA, 1:5000), DMPO (gif of Ronald Mason, National Institute for Environmental Health Sciences, Research Triangle Park, NC, USA), SOD2 (#06-984, Upstate EMD Millipore, Burlington, MA, USA, 1:1000). β-actin (#ab8227, abcam, Cambridge, MA, USA) and calnexin (#ADI-SPA-860-F, Enzo, Farming- dale, NY, USA) were used as loading controls.

Cell viability assays and ­H2O2 treatment. To test for a diference in baseline viability, MEFs were seeded in 6 ­cm2 plates. Total cells were collected for up to 5 days post seeding. Cells were mixed with trypan blue. Cells excluding trypan blue were denoted as live with the percentage of live cells divided total used to calculate viability. Cells treated with ­H2O2 received treatment for 24 h at a 500 µM concentration.

NADP+/NADPH ratio determination. Reagents from Abcam (catalog # ab65349) were used according to manufacturer protocols as described ­previously67. Tree million WT and Rev-erbα KO cells were utilized for each biological measurement. DNA was sheared via syringing prior to ­NADP+ decomposition. Optical plates were read at 1 h following addition of assay developer. For individual total species and NADPH quantities, assay readings were normalized to protein amounts quantifed via a BCA assay.

Statistical analysis. Experiments were repeated biologically three times, with three technical triplicates for quantitative PCR and functional assays. Seahorse assays contained up to nine technical replicates each. Error bars are reported as the standard error of the mean. Graph Pad Prism 9 sofware was used for t-test and ANOVA analysis. Te unpaired t test was used for detecting statistical signifcance of the diferences between means of two groups if the data are normally distributed. Welch-corrected t test was used if the data are not normally distributed. Te statistical signifcance of the diferences among multiple groups was evaluated by using one-way ANOVA for overall signifcance, followed by Tukey–Kramer test. Statistical signifcance was considered when the p value was less than 0.05.

Received: 5 September 2020; Accepted: 19 May 2021

References 1. Fontaine, C. et al. Te orphan nuclear receptor Rev-Erbalpha is a peroxisome proliferator-activated receptor (PPAR) gamma target gene and promotes PPARgamma-induced adipocyte diferentiation. J. Biol. Chem. 278, 37672–37680 (2003). 2. Preitner, N. et al. Te orphan nuclear receptor REV-ERBalpha controls circadian transcription within the positive limb of the mammalian circadian oscillator. Cell 110, 251–260 (2002). 3. Raspe, E. et al. Identifcation of Rev-erbα as a physiological repressor of apoC-III gene transcription. J. Lipid Res. 43, 2172–2179 (2002). 4. Mohwak, J. A., Green, C. B. & Takahashi, J. S. Central and peripheral circadian clocks in mammals. Annu. Rev. Neurosci. 2012, 445–462 (2012). 5. Zhang, Y. et al. GENE REGULATION. Discrete functions of nuclear receptor Rev-erbalpha couple metabolism to the clock. Science 348, 1488–1492. https://​doi.​org/​10.​1126/​scien​ce.​aab30​21 (2015). 6. Duez, H. & Staels, B. Rev-erb alpha gives a time cue to metabolism. FEBS Lett. 582, 19–25 (2008). 7. Solt, L. A. et al. Regulation of circadian behavior and metabolism by synthetic REV-ERB agonists. Nature 485, 62–68 (2012). 8. Amador, A. et al. Distinct roles for REV-ERBα and REV-ERBβ in oxidative capacity and mitochondrial biogenesis in skeletal muscle. Plos One 13, e0196787 (2018). 9. Woldt, E. et al. Rev-erb-α modulates skeletal muscle oxidative capacity by regulating mitochondrial biogenesis and autophagy. Nat. Med. 19, 1039–1046 (2013). 10. Dyar, K. A. et al. Transcriptional programming of lipid and metabolism by the skeletal muscle circadian clock. PLOS Biology 16, e2005886 (2018). 11. Zhang, Y. et al. HNF6 and Rev-erbα integrate hepatic lipid metabolism by overlapping and distinct transcriptional mechanisms. Genes Dev. 30, 1636–1644 (2016). 12. Ikeda, R. et al. REV-ERBα and REV-ERBβ function as key factors regulating mammalian circadian output. Sci. Rep. 9, 10171 (2019). 13. Sengupta, S. et al. Te circadian gene Rev-erbα improves cellular bioenergetics and provides preconditioning for protection against oxidative stress. Free Radical Biol. Med. 93, 177–189 (2016). 14. Patra, K. C. & Hay, N. Te pentose phosphate pathway and cancer. Trends Biomed. Sci. 39, 347–354 (2014). 15. Boros, L. G. et al. Oxythiamine and dehydroepiandrosterone inhibit the nonoxidative synthesis of ribose and tumor cell prolifera- tion. Can. Res. 57, 4242–4248 (1997). 16. Popovici, T., Berwald-Netter, Y., Vibert, M., Kahn, A. & Skala, H. Localization of aldolase C mRNA in brain cells. FEBS Lett. 268, 189–193 (1990). 17. Tompson, R. J., Kyonch, P. A. M. & Willson, V. J. C. Cellular localization of adolase C subunits in the human brain. Brain Res. 232, 489–493 (1981). 18. Han, Z., Zhong, L., Srivastava, A. & Stacpoole, P. W. Pyruvate dehydrogenase complex defciency caused by ubiquitination and proteasome-mediated degradation of the E1β subunit. J. Biol. Chem. 283, 237–243 (2008).

Scientifc Reports | (2021) 11:12356 | https://doi.org/10.1038/s41598-021-91516-5 10 Vol:.(1234567890) www.nature.com/scientificreports/

19. Fang, R., Nixon, P. F. & Duggelby, R. G. Identifcation of the catalytic glutamate in the E1 component of human pyruvate dehy- drogenase. FEBS Lett. 437, 273–277 (1998). 20. Yin, L., Wang, J., Klein, P. S. & Lazar, M. A. Nuclear Receptor Rev-erba is a critical lithium sensing component of the circaidian clock. Science 331, 1002–1005 (2006). 21. Lee, C. R., Park, Y. H., Min, H., Kim, Y. R. & Seok, Y. J. Determination of protein phosphorylation by polyacrylamide gel electro- phoresis. J. Microbiol. 57, 93–100 (2019). 22. Wegener, A. D. & Jones, L. R. Phosphorylation-induced mobility shif in phospholamban in sodium dodecyl sulfate-polyacrylamide gels. J. Biol. Chem. 259, 1834–1841 (1984). 23. Benito, A. et al. Glucose-6-phosphate dehydrogenase and transketolase modulate breast cancer cell metabolic reprogramming and correlate with poor patient outcome. Oncotarget 8, 106693–106706 (2017). 24. Qiu, Z. et al. MicroRNA-124 reduces the pentose phosphate pathway and proliferation by targeting PRPS1 and RPIA mRNAs in human cells. Gastroenterology 149, 1587–1598 (2015). 25. Ciou, S. C. et al. Ribose-5-phosphate isomerase A regulates hepatocarcinogenesis via PP2A and ERK signaling. Int. J. Cancer 137, 104–115 (2015). 26. Boros, L. G. et al. Oxythiamine and dehydroepiandrosterone inhibit the nonoxidative synthesis of ribose and tumor cell prolifera- tion. Can. Res. 57, 4242–4248 (1997). 27. Rais, B. et al. Oxythiamine and dehydroepiandrosterone induce a G1 phase cycle arrest in Ehrlich’s tumor cells through inhibition of the pentose cycle. FEBS Lett. 456, 113–118 (1999). 28. Wang, J. et al. Inhibition of transketolase by oxythiamine altered dynamics of protein signals in pancreatic cancer cells. Exp. Hematol. Oncol. 2, 18 (2013). 29. Af, R. L., Zhang, F. W. & Gius, D. Evaluation of 2-deoxy-d-glucose as a chemotherapeutic agent: mechanism of cell death. Br. J. Cancer 87, 805–812 (2002). 30. Wick, A. R., Drury, D. R., Nakada, H. I. & Wolfe, J. B. Localization of the primary metabolic block produced by 2-deoxyglucose. J. Biol. Chem. 224, 963–969 (1956). 31. Yang, G. et al. Oxidative stress and infammation modulate Rev-erbα signaling in the neonatal lung and afect circadian rhythmic- ity. Antioxid. Signal. 21, 17–32 (2014). 32. Janzen, E. G., Jandristis, L. T., Shetty, R. V., Haire, D. L. & Hilborn, J. W. Synthesis and purifcation of 5,5-dimethyl-1-pyrroline- N-oxide for biological applications. Chem. Biol. Interact. 80, 167–172 (1989). 33. Ranguelova, K. & Mason, R. P. Te fdelity of spin trapping with DMPO in biological systems. Magn. Reson. Chem. 49, 152–158 (2011). 34. Frankel, D., Mehindate, K. & Schipper, H. M. Role of heme oxygenase-1 in the regulation of manganese superoxide dismutase gene expression in oxidatively-challenged astroglia. J. Cell. Physiol. 185, 80–68 (2000). 35. Gregory, E. M., Goscin, S. A. & Fridovich, I. Superoxide dismutase and oxygen toxicity in a eukaryote. J. Bacteriol. 117, 456–460 (1974). 36. Vile, G. F., Basu-Modak, S., Waltner, C. & Tyrrell, R. M. Heme oxygenase 1 mediates an adaptive response to oxidative stress in human skin fbroblasts. Proc. Natl. Acad. Sci. USA 91, 2607–2610 (1994). 37. Dierichkx, P. et al. SR9009 has REV-ERB–independent efects on cell proliferation and metabolism. Proc. Natl. Acad Sci. USA 116, 12147–12152 (2019). 38. Zhang, T. et al. REV-ERBα regulates CYP7A1 through repression of liver receptor homolog-1. Drug Metab. Dispos. 46, 248–258 (2018). 39. Zhang, Y. et al. HNF6 and Rev-erbalpha integrate hepatic lipid metabolism by overlapping and distinct transcriptional mechanisms. Genes Dev. 30, 1636–1644. https://​doi.​org/​10.​1101/​gad.​281972.​116 (2016). 40. Toledo, M. et al. Autophagy regulates the liver clock and glucose metabolism by degrading CRY1. Cell Metab. 28, 268–281 (2018). 41. Yuan, X., Dong, D., Li, Z. & Wu, B. Rev-erbα activation down-regulates hepatic Pck1 enzyme to lower plasma glucose in mice. Pharmacol. Res. 141, 310–318 (2019). 42. Delezie, J. et al. Te nuclear receptor REV-ERBa is required for the daily balance of carbohydrate and lipid metabolism. FASEB J. 26, 3321–3335 (2012). 43. Gibbs, J. E. et al. Te nuclear receptor REV-ERBα mediates circadian regulation of innate immunity through selective regulation of infammatory cytokines. Proc. Natl. Acad. Sci. U.S.A 109, 582–587 (2012). 44. Amir, M. et al. REV-ERBα regulates TH17 cell development and autoimmunity. Cell Rep. 25, 3733–3749 (2018). 45. Chang, C. et al. Te nuclear receptor REV-ERBα modulates T17 cell-mediated autoimmune disease. Proc. Natl. Acad. Sci. USA 116, 18528–18536 (2019). 46. Fuhr, L. et al. Te circadian clock regulates metabolic phenotype rewiring via HKDC1 and modulates tumor progression and drug response in colorectal cancer. EBioMedicine 33, 105–121 (2018). 47. Hunter, A. L. et al. Nuclear receptor REVERBalpha is a state-dependent regulator of liver energy metabolism. Proc. Natl. Acad. Sci. USA 117, 25869–25879. https://​doi.​org/​10.​1073/​pnas.​20053​30117 (2020). 48. Balsalobre, A., Damiola, F. & Schibler, U. A serum shock induces circadian gene expression in mammalian tissue culture cells. Cell 93, 929–937. https://​doi.​org/​10.​1016/​s0092-​8674(00)​81199-x (1998). 49. Ling, L. et al. Transcriptional regulation of the Warburg efect in cancer by SIX1. Cancer Cell 33, 368–385 (2018). 50. Weber, G. Enzymology of cancer cells. N. Engl. J. Med. 10, 541–551 (1977). 51. Warburg, O. On the origin of cancer cells. Science 123, 309–314 (1956). 52. Sulli, G. et al. Pharmacological activation of REV-ERBs is lethal in cancer and oncogene-induced senescence. Nature 553, 351–355 (2018). 53. Burgermeister, E. et al. Aryl hydrocarbon receptor nuclear translocator-like (ARNTL/BMAL1) is associated with bevacizumab resistance in colorectal cancer via regulation of vascular endothelial growth factor A. EBioMedicine 45, 139–154 (2019). 54. Wagner, P. M., Monjes, N. M. & Guido, M. E. Chemotherapeutic Efect of SR9009, a REV-ERB Agonist, on the Human Glioblastoma T98G Cells. ASN Neuruo 11, 1–14 (2019). 55. Na, H. et al. High expression of NR1D1 is associated with good prognosis in triple-negative breast cancer patients treated with chemotherapy. Breast Cancer Res. 21, 217 (2019). 56. Ka, N., Na, T. & Lee, M. NR1D1 enhances oxidative DNA damage by inhibiting PARP1 activity. Mol. Cell. Endocrinol. 454, 87–92 (2017). 57. Tao, L. et al. Rev-erbalpha inhibits proliferation by reducing glycolytic fux and pentose phosphate pathway in human gastric cancer cells. Oncogenesis 8, 57. https://​doi.​org/​10.​1038/​s41389-​019-​0168-5 (2019). 58. Chu, G., Zhou, Q., Hu, Y., Shengjie, S. & Yang, G. Rev-erb inhibits proliferation and promotes of preadipocytes through the agonist GSK4112. Int. J. Mol. Sci. 20, 4524 (2019). 59. Bustamante, E. & Pedersen, P. L. High aerobic glycolysis of rat hepatoma cells in culture: role of mitochondrial hexokinase. Proc. Natl. Acad. Sci. U.S.A. 74, 3735–3739 (1977). 60. Patra, K. C. et al. Hexokinase 2 is required for tumor initiation and maintenance and its systemic deletion is therapeutic in mouse models of cancer. Cancer Cell 24, 213–228 (2013). 61. Wang, L. et al. Hexokinase 2-mediated warburg efect is required for PTEN- and -defciency-driven prostate cancer growth. Cell Rep. 8, 1461–1474 (2014).

Scientifc Reports | (2021) 11:12356 | https://doi.org/10.1038/s41598-021-91516-5 11 Vol.:(0123456789) www.nature.com/scientificreports/

62. Taro, H. et al. Tyrosine phosphorylation inhibits PKM2 to promote the Warburg efect and tumor growth. Sci. Signal. 2, 73 (2010). 63. Christof, H. R. et al. Te M2 splice isoform of pyruvate kinase is important for cancer metabolism and tumour growth. Nature 452, 230–233 (2008). 64. Wang, J. et al. Te platelet isoform of phosphofructokinase contributes to metabolic reprogramming and maintains cell prolifera- tion in clear cell renal cell carcinoma. Oncotarget 7, 27142–27157 (2016). 65. Li, L. et al. TAp73-induced phosphofructokinase-1 transcription promotes the Warburg efect and enhances cell proliferation. Nat. Commun. 9, 4683 (2018). 66. Yao, H. et al. Fatty acid oxidation protects against hyperoxia-induced endothelial cell apoptosis and lung injury in neonatal mice. Am. J. Respir. Cell Mol. Biol. 60, 667–677. https://​doi.​org/​10.​1165/​rcmb.​2018-​0335OC (2019). 67. Gong, J. et al. Te pentose phosphate pathway mediates hyperoxia-induced lung vascular dysgenesis and alveolar simplifcation in neonates. JCI Insight https://​doi.​org/​10.​1172/​jci.​insig​ht.​137594 (2021). Acknowledgements Tis work was supported by the Warren Alpert Foundation to PAD, the Institutional Development Award (IDeA) from the NIGMS of NIH under Grant #P20GM103652 and the Falk Medical Research Trust Catalyst Award to HY. We thank Dr. Mitchell Lazar for providing the Rev-erb MEF cell lines and Abigail Peterson for aiding in the analysis of the microarray data. Author contributions Conception and Design: S.P.G., H.Y., P.A.D.; Data acquisition and analysis: S.P.G., H.Y., S.R., H.M., J.C.; Data Interpretation: S.P.G., H.Y., P.A.D.; Drafing of the manuscript: S.P.G.; Revision of the manuscript: H.Y., P.A.D.

Competing interests Te authors declare no competing interests. Additional information Supplementary Information Te online version contains supplementary material available at https://​doi.​org/​ 10.​1038/​s41598-​021-​91516-5. Correspondence and requests for materials should be addressed to P.A.D. Reprints and permissions information is available at www.nature.com/reprints. Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional afliations. Open Access Tis article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. Te images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://​creat​iveco​mmons.​org/​licen​ses/​by/4.​0/.

© Te Author(s) 2021

Scientifc Reports | (2021) 11:12356 | https://doi.org/10.1038/s41598-021-91516-5 12 Vol:.(1234567890)